With the support of preceding grants, the PI succeeded in characterizing almost all iron-sulfur clusters and protein-bound ubiquinone in complex I. Now, we are ready to launch a new approach in this field. Recently Sazanov's group successfully determined the X-ray crystallographic structure (3.3A resolution) of the peripheral part of complex I isolated from thermophilic bacteria. Their results are fully consistent with the structures predicted by the Pi's EPR studies as described in this application. Specifically, X-ray data indicated that iron-sulfur cluster N2, which directly donates electrons to protein-associated ubiquinone (designated SQNf), is located only 8A from the membrane domain interface. They also reported that the cluster N2 is located close to a nearby short channel leading to ubiquinone. All of their data are consistent with the Pi's proposal of the N2"-"SQNf distance of 12A. The PI found that the SQNf signal disappeared upon addition of uncouplers. This unique property cannot be explained through the classic chemiosmotic loop mechanism. In order to prove that the SQNf signal is sensitive to the membrane potential, we prepared proteoliposomes reconstituted from complex I. We succeeded in demonstrating that the signal is enhanced with an inside-positive membrane potential which was induced by the K-valinomycin method. This system will enable us to apply various advanced magnetic resonance techniques, high frequency EPR and newly developed Relaxation filtered hyperfine spectroscopy, combined as REFINE-ENDOR or REFINE-HYSCORE, in order to detect possible conformational changes that the gating semiquinone may undergo. We will also study the proton pump mechanism within the membrane, using modern molecular biological techniques on bacterial complex I system. Many mitochondria-linked genetic diseases, Parkinson's disease, apoptosis and aging are related with complex I defect(s) and/or generation of reactive oxygen species (ROS) from complex I. Complex I is the major site of superoxide generation in brain and heart mitochondria, but no consensus has been attained regarding the generation site(s). Thus, to identify the generation site is vital for better understanding these problems, and for developing therapeutic managements. Now, we have established an entirely new, powerful tool for this study. We measure signals of semiquinone, superoxide-spin adduct, and iron-sulfur signals from the same EPR sample. Using this new strategy, we found that both cluster N2 and flavin can generate superoxide. We will solidify this exciting finding and study its relevance to complex I function under physiological and pathological conditions.